README: Update training plots with final winning results (93% compliance, 0->4.5 reward)

Remove broken auto-generated plots in favor of live Kaggle training evidence

Update README with requirements alignment, Colab/Kaggle links, and live Kaggle training evidence

Update latest changes

Merge origin

chore: exclude binary files to allow push to hf space

Hackathon Final Submission: PettingZoo multi-agent arch, GRPO training, docs

Final cleanup, configuration, and rebranding to QuantHive for HF deployment